Yagna Thakkar
@yagnathakkar
Machine Learning Intern at Hex Wireless PVT LTD
Anand, Gujarat, India
Yagna Thakkar is a skilled ML/CV engineer with experience in developing deep learning solutions. Proficient in frameworks like TensorFlow, PyTorch, and Keras, with expertise in Computer Vision (YOLO, OpenCV) and NLP (LSTM). Proven ability to work on real-time systems, including speaker diarization and object detection, complemented by strong data analysis skills using Power BI and Python.
Experience
Machine Learning Intern
Hex Wireless PVT LTD
Work on Real-Time Speaker Diarization System pipeline. Extracting Human Voice Features (MFCC), Audio Analysis using PyAudio and Matplotlib. Used Pydub for Audio Manipulation. Use webrtcvad package for Voice Activity Detection and DeepSpeech for Transcription. Pyannote library to diarize entire based on Speech Brain. Dashboard to analyze audio and generate speaker separated audio segments.
Computer Vision Intern
Artenal
To build A.I. powered and Deep Learning based algorithms for harvesting fruits more efficiently. Collect and Preprocess Custom Dataset. Filtering of Data, Data analysis, labelling, model training and hyper-parameter tuning. Model Ensembling. Hyper-parameter evolution to find optimal values. Evaluate YOLOv5 performance and Visualize training data. Successfully detected apples and grapes for harvest. Managed the Computer Vision team of 5 Members.
Machine Learning Intern
GRROOM
To build an object detection model to detect whether or not the person is wearing a mask. Collect and Preprocess custom data, went through various YOLOV3 documentation as a part of the process. Implemented darknet for GPU computation. Prepared data for training and yolov3 testing. Used CFG files in OpenCV script to predict if mask is worn or not. Successfully detected mask with an accuracy of 80%.
Education
SARDAR VALLABHBHAI PATEL INSTITUTE OF TECHNOLOGY - SVIT (GTU)
B.E. (E&C)
E&C
Final Year
V & C PATEL ENGLISH SCHOOL
Class XII
PCM